Internal Document
Category: Meta Size: 33.5 KB Lines: 325 Commit: 36ca060 Modified: Apr 26, 2026

Agent9 Marketing Plan — Project Log

Save-as-we-go log of every meaningful step in the marketing-plan build. Updated incrementally and committed at each step so nothing is ever in flight without disk persistence.

Owner Brief (Apr 25 2026, 22:22 ET)

Owner directive (Telegram message 21048 + 21060):

Owner's six framing answers (Apr 26 02:22 ET):

  1. Showing model — portfolio of 3 options with primary recommendation.
  2. Lead-scraping legality — frank version. Include big firms (Sibcy Cline, Coldwell Banker, KW, EXP, Re/Max).
  3. Geography — Ohio for now; framework first then national expansion.
  4. Buyer-side — defer; focus on seller acquisition via existing-listings prospecting.
  5. Brand — keep agent9/9enterprises identity unless team finds a liability.
  6. Polish — same as agent9 vision page (Owner-only review-grade).

Stage 1 — Research (4 parallel dispatches, in flight)

ID Agent Topic Output target Status
S1-A UNO Scraping legality matrix + workarounds (Zillow, Realtor, Houzeo, FB, Craigslist, MLS, county auditors, yard-sign capture, expired-listing precedent, tortious interference risk on realtor-listed sellers) research/scraping-legality-matrix.md DONE (210 lines, 2,975 words)
S1-B SCOUT Ohio big-firm landscape (Sibcy Cline, CB, KW, EXP, Re/Max) — competitor / data source / agent-supply research/ohio-firm-landscape.md DONE (201 lines, 2,375 words)
S1-C UNO Showing model deep-dive — 3 options trade-off matrix + comp model for option C research/showing-model-deep-dive.md DONE (209 lines, 2,850 words)
S1-D SCOUT Ohio outreach legal compliance (TCPA / CAN-SPAM / ORC 4735 / DNC) — channel × seller-segment research/ohio-outreach-compliance.md DONE (228 lines, 2,689 words)

Stage 1 findings landed so far

S1-A scraping legality (UNO):

S1-B Ohio firm landscape (SCOUT):

S1-C showing-model deep-dive (UNO):

S1-D Ohio outreach compliance (SCOUT):

Stage-1 → Stage-2 takeaways for the strategy synthesis

  1. The marketing plan MUST bake the tortious-interference copy framing into every piece of seller-with-realtor outreach. Two independent arms surfaced this. "Cancel and switch" wording = exposure. "Alongside / when ready / when expires" wording = safe.
  2. Showing model defaults to Option B (self-tour ID-verified lockbox). Option A breaks the brand thesis. Option C is the upgrade path post-attorney-review.
  3. Acquisition channel mix: Direct mail (lead) + Email (volume amplifier) + County-auditor public records as the lead-source backbone. Cold SMS killed. Cold voice gated behind $50k bond + AG registration — surface as a Year-2 unlock, not Year-1 default.
  4. Agent supply pool for Option C = EXP first, KW pre-cap second, RE/MAX desk-fee third. Showami already pays $32 avg/showing — Agent9 must price ≥ $40 to compete for that supply.
  5. Real Ohio competitor = Houzeo at flat-fee tier; not the big brands. No major brand has entered the flat-fee lane in Ohio. Agent9 differentiates above MLS-only with AI comms + showing network + txn mgmt.

ETA: 25-30 min per arm. Stage 2 cannot start until all 4 land.

Stage 2 — Strategy synthesis (PRESS) — DONE

File: marketing-plan.md (371 lines, 5,521 words). 11 sections covering Exec Summary, Positioning, Two-Segment Targeting, Lead Sources, Outreach Channels, Showing Model, Showing-Agent Comp, Ohio Launch Sequence, Year-2 Expansion, What We're Not Doing, Risks/Open Questions.

PRESS-flagged honest gaps the Owner needs to know about:

  1. Ohio-specific expired-listing count not available in Stage 1 — Segment A TAM is directional, not exact.
  2. Michigan FSBO rate cited as expansion candidate but not confirmed — verify before committing.
  3. Lockbox vendor selection open (Igloohome flagged as lead, no contract terms confirmed).
  4. Brand variants assume current agent9 identity — re-skin if Stage 3 CANVAS recommends alternative.
  5. Option C (showing agent) attorney review on ORC 4735 is a HARD gate — memo before build sprint, not after.

Stage 3 — Brand assessment (CANVAS) — DONE

File: research/brand-assessment.md (141 lines, 2,893 words). Verdict: KEEP-WITH-MODIFICATIONS for the marketing context. Vision page (the agent9 deliverable Owner reviewed tonight) gets ZERO changes.

Three modifications for the marketing page only:

  1. Restrain orange. Deploy #FB4F14 only on three conversion moments — CTA button, wordmark 9, pricing callout. Currently every accent is orange at full saturation; reads energetic but not fiduciary for the 55+ seller-with-realtor segment.
  2. Add trust signals block. Three testimonials (or "concept testimonials" tagged as such pre-launch), Ohio license disclosure inline, physical address, review-badge placeholder. Category expectation set by Houzeo's Trustpilot 4.9 / 4,000+ reviews. CRO research suggests up to 42% conversion lift from trust signal addition.
  3. Wordmark sub-identifier. Add Ohio Real Estate · Sell without a Realtor at 11px under the nav logo on marketing page only. Wordmark does not self-identify category on a direct mail envelope or yard sign without it.

Stage 4 implementation guidance is embedded in the file (specific hex values, CSS class suggestions, font import, section placement) so HTML assembly can implement directly.

Honest gap: no Ohio-specific FSBO color-preference A/B data exists. Orange-risk finding is from category-wide color psychology, not Ohio specific. If Owner wants proof before committing to the orange-restraint mod, alt path is ship full-orange + trust signals first, then A/B test orange intensity post-launch.

Stage 4 — HTML assembly (CANVAS) — DONE

File: agent9-marketing-plan.html (2,502 lines, 97KB). 13 sections: hero, segments, positioning, lead-sources, channels, showing, comp-model, launch, expansion, not-doing, risks, trust, closing.

CANVAS self-gate: 0px iPhone horizontal overflow at 390x844, 13/13 sections render, sub-identifier 11px, trust block visible, draft badge visible. 5 in-flight orange-restraint fixes applied during gate (volume table highlight, showing primary badge, comp-table self-row, ladder-card category-leader ARR value, nav-sub responsive clamp). Final orange usage: 8 hits, all on conversion moments (wordmark 9 x2, nav CTA, badge dot, btn-primary x2, $999 price, closing pricing callout). Zero decorative orange.

Screenshots committed at agent9/marketing/screenshots/:

CANVAS-flagged honest gap: Google Fonts CDN console warning on first load (network only, not structural; fonts loaded successfully in both screenshots).

Deployment

Mirrored to public/agent9-marketing/index.html and dist/agent9-marketing/index.html. _routes.json + _redirects updated for both. CF Pages auto-deploy fires on push.

Live URL: https://9enterprises.ai/agent9-marketing/ (after auto-deploy completes — 1-2 min from push).

Save discipline status

Every artifact landed within 5 min of agent return. Five commits across the marketing-plan v1 build:

Round 2 — Owner review feedback (Apr 26 ~00:23 ET)

Owner reviewed v1 and sent refinements (Telegram 21071):

Owner also asked: "What are our current limitations and needs in relation to being able to fully service all of the aspects of reliable communication what is the honest scope of the infrastructure build out?" + "is the next natural step a functional demo website?"

Round 2 — 4 parallel reviews dispatched

ID Agent Topic Output target Status
R2-A UNO Legal review of buyer-channel-inquiry approach (Ohio CSPA + tortious interference re-frame + Zillow/Realtor TOS on inquiry channel) marketing/research/v2-buyer-channel-inquiry-legality.md DONE (216 lines, 3,400 words)
R2-B SCOUT Brutal-honesty social/viral marketing read for FSBO real estate marketing/research/v2-social-viral-honest-read.md DONE (288 lines, 27KB)
R2-C UNO Infrastructure capability honest scope — what 9enterprises has, what Agent9 needs, gap map, Q3 MVP shape, demo question infrastructure/v2-honest-scope-assessment.md DONE (3,717 words, 23KB)
R2-D Agent9 GM Product synthesis — hybrid showing rules, lockbox economics, agent network, dedicated-agent subscription, mobile app verdict, demo MVP shape marketing/research/v2-product-synthesis.md DONE (318 lines, 4,402 words)

R2-C key findings (UNO infrastructure scope)

Two HARD BLOCKERS that cannot be built around:

  1. Ohio licensed broker for MLS submission (partnership required — ORC 4735)
  2. Ohio real estate attorney to clear the unlicensed-broker question on Agent9 acting as payment-facilitator-for-prospect-procurement

Until both are in motion this week, no code written for the demo matters.

Q3 2026 verdict:

UNO honest gaps: Igloohome API pricing + lead time not confirmed (need dev account); Twilio A2P 10DLC registration status not verified (check Locker / Twilio dashboard before SMS).

R2-A key findings (UNO buyer-channel-inquiry legality)

VERDICT: Framing A (Owner's verbatim) is LEGALLY TOXIC. Framing B is usable but NOT on Zillow at volume.

Zillow is NOT the channel at any volume:

Viable channels for Framing B at volume: Craigslist FSBO, ForSaleByOwner.com, Houzeo (partnership), direct FSBO outreach via county auditor data. Cross-references S1-A's "BUILD AROUND" tier exactly.

NAR Code of Ethics Article 16 explicitly applies only to REALTORS® members — does not reach Agent9. The realtor-industry restriction on contacting active-listing sellers is a member ethics rule, not a legal constraint on tech companies.

Tortious interference still in play even on solicited inquiry channel if the copy pitches breach. Same content rules from S1-A apply: "when your listing expires" not "cancel now."

Required pre-launch: $400-$600 Ohio attorney memo pre-clearing the boilerplate copy template on CSPA + ORC 4735.02 unlicensed-practice. Same memo covers the Option C ORC 4735 question from S1-C.

UNO honest gap: Zillow TOS page returned 403 to automated fetch; prohibited-use language is confirmed from multiple indexed secondary sources but should be manually verified at zillow.com/z/corp/terms/ before any campaign decision.

R2-B key findings (SCOUT social/viral honest read)

On "viral": ZERO FSBO platforms have grown via organic social. Not one. Houzeo grew on SEO (200K+ organic searches/month) + Google reviews (8,500+) + TrustPilot reviews (2,000+) + paid Google search ads. That is the real playbook. "Viral" does not apply to a once-per-decade high-anxiety transaction.

Channel verdicts (calibrated):

Homepage convergent pattern (validates Owner ask): hero savings claim + 3-step "how it works" + 60-90s demo video + transparent pricing no asterisks + FAQ on 5 core legal/logistical objections + trust signals + compliance footer. Commission-savings calculator widget = the ONE proven viral lead-capture tool in this category. Build at launch.

Mobile app verdict: Do NOT build native iOS/Android for Q3 2026. Beycome processes transactions at volume with no app. Houzeo built native AFTER establishing the platform. PWA-capable mobile-optimized web is the right MVP call. Native = Year-2.

Year-1 social budget: $1,800-$4,600/mo total. YouTube production is the majority. Everything else near-zero. Real Year-1 investment = direct mail + SEO content + review-collection workflow built into product — those three are what Houzeo actually used to win.

R2-D key findings (Agent9 GM product synthesis)

Showing-model conflict cases — RESOLVED:

Lockbox economics: retail-parity pricing $149 sell vs $89 Igloohome wholesale = $45-54 net margin/unit after Stripe + CS overhead. Floor matches Owner's $35-45 estimate; ceiling $54-74 with config fee. HONEST FLAG: same hardware on Amazon. Either price at parity + position on convenience/pre-configuration, or charge a setup fee separately. Sellers will comparison-shop.

Showing-agent economics revised: Owner's $50/$10/$40 split has THIN $10 retention. Revised to $60/$15-20/$40-45. EXP + KW pre-cap = first recruitment targets. Don't advertise Option C until 25 verified agents onboarded in Columbus + Cincinnati.

LEGAL FLAG — hardest stop in the entire plan: ORC 4735's "real estate broker" definition is broad enough to potentially cover Agent9's Option C fee-facilitation role. The "$400-800 attorney review then ship" path may be optimistic. Attorney may come back recommending a licensed Ohio managing broker on staff or in formal partnership before dispatch. Managing broker partnership takes time to structure + revenue sharing. Plan for 2-3 months to resolve properly, not 2-3 weeks.

Dedicated showing agent subscription — MATH PROBLEM: Owner's $299 flat for 10 showings = $29.90/showing implied — BELOW agent floor of $40. Three resolution paths in doc; cleanest = monthly retainer model: $199/mo from buyer, $120/mo retainer to dedicated agent, Agent9 keeps $79. Avoids per-showing counting disputes. This pricing decision must be made BEFORE the dedicated subscription is built — it drives the agent contract structure.

Buyer LTV reality: dedicated subscription LTV ≈ $158/buyer vs à la carte ≈ $200/buyer at first pass. Subscription ONLY makes economic sense if it increases tour volume or offer-conversion rates — not as a standalone margin play. Worth modeling once we have first-cohort data.

Mobile app verdict: Native iOS + Android NOT realistic for Q3 2026 on any reasonable budget. PWA is the call. Only real limitation vs native = background notifications, which aren't required for Agent9's core use case at launch.

Homepage convergent pattern: savings calculator FIRST, then 3-step how-it-works, then specific-dollar testimonials, then objection FAQ, then final CTA. 5 sections specified. Stage 3 trust signals confirmed BUT specificity matters — generic security badges don't convert. Real dollar amounts from real Ohio sellers do. Pre-launch trust signals = "Concept testimonial — pre-launch" tagging from Stage 3, then replace with real once first 5 transactions close.

Demo MVP verdict: YES, build it. 8-week sprint achievable using the existing 9E Solution comms layer. Doc distinguishes real-functional-from-day-1 vs wizard-of-oz back-filled. Demo lives at agent9.9enterprises.ai for internal sprint, then move to dedicated domain before first real seller onboards.

Round 2 — Synthesis: critical-path decisions for Owner

Convergent verdicts (all 4 reports agree):

  1. Framing A is dead. "Pretextual buyer then pivot" violates Ohio CSPA + FTC. Use Framing B (transparent Agent9 identification in sentence one) — but NOT on Zillow at any volume (TOS + Imperva detection trip ≤10 inquiries/day).
  2. "Viral" is dead for this category. No FSBO platform has grown via organic social. The real playbook = SEO + reviews + Google paid + commission-savings calculator widget. YouTube is the one social channel that works (compounds with SEO).
  3. Mobile = PWA, not native. Three reports converge.
  4. Q3 2026 = Demo achievable, full production NOT achievable with current 2-person team.

Two HARD BLOCKERS that must start moving this week:

  1. Ohio licensed broker partnership for MLS submission AND possibly mandated for Option C fee-facilitation
  2. Ohio real estate attorney memo ($400-800 confirmed cost) that may recommend a managing broker on staff before Option C ships

Decision asks for Owner:

Round 2.5 — Owner MLS-skip pivot proposal (Apr 26 ~01:40 ET)

Owner asked: skip MLS entirely, list on Zillow + Trulia (free), drive traffic to own site. Goal: alleviate broker-licensure concerns. Dispatched UNO for honest research.

R2.5 UNO key findings (marketing/research/v2-mls-skip-analysis.md, 2,100 words)

VERDICT: MLS-skip does NOT deliver the legal relief Owner expects, and INTRODUCES a new operational blocker.

  1. Zillow FSBO IS free — Owner's claim correct. No fees, no expiration. Premium placement optional.
  2. Trulia IS auto-syndicated from Zillow — Owner's claim correct. One workflow, both platforms, no extra cost.
  3. ORC 4735 still triggers without MLS. Statute lists 6+ broker-defining activities (list, advertise, procure prospects, assist negotiations, hold-out as engaged, receive compensation for any of the above). MLS submission is just ONE. The compensation-for-listing-and-procuring structure that Agent9 charges $999 for hits the statute regardless of MLS membership. Removing MLS removes one trigger; six others remain fully intact.
  4. Buyer pool collapses ~88% → ~15-20% without MLS. 88% of Ohio buyers (NAR 2025) use a buyer's agent. Buyer's agents work from MLS IDX feeds; they do not browse Zillow FSBO. Zillow itself confirmed in REX v. Zillow litigation that FSBO listings get 80-85% fewer page views than MLS-listed properties. Each listing takes longer to sell, weaker offers, higher abandonment.
  5. NEW operational blocker: Zillow's Listings Quality Policy requires the property owner to post. Zillow can require certificate of title to verify ownership. Agent9 cannot post listings claiming to own other people's properties. This means "Agent9 lists and manages the property on those platforms" does not work on Zillow's platform as currently structured. Each seller would need to self-post — fundamentally changes Agent9's service model from "we handle it" to "we coach you through doing it yourself."
  6. Houzeo + Beycome counterfactual: both maintain Ohio broker licenses (Houzeo BRKP.2022001015, Beycome REC.2025001306). They DO submit to MLS even though they could Zillow-only. They hold these licenses because the activities trigger ORC 4735 regardless of MLS. If Zillow-only worked legally and commercially, the entire FSBO category would already do it. None do.
  7. Attorney memo scope expands $200-$400 to cover both scenarios in single engagement.
  8. Recommendation: MLS-in with licensed Ohio broker partner. $50-$150/listing. Broker submits to MLS, MLS syndicates to Zillow/Trulia automatically. Solves the Zillow listing management problem. Puts licensed pro in the chain for ORC 4735 compliance. Full 88% buyer pool.

Honest gaps from UNO: Zillow TOS page returned 403 (owner-must-post confirmed via secondary sources); "500x more visitors" claim unverified (REX v. Zillow 80-85% figure is better-sourced); no Ohio Division of Real Estate formal opinion on AI platforms exists — attorney memo is the only path to certainty.

Round 2.6 — Owner refined pivot: seller-self-post + AI comms-layer (Apr 26 ~02:06 ET)

Owner refined: seller signs up → Agent9 chatbot walks them through Zillow listing self-post (under "9 minutes"), seller posts their own listing, Agent9 then operates as "communication layer between them and potential buyers" with seller's consent. Question: does this clear ORC 4735?

R2.6 UNO key findings (marketing/research/v2-seller-self-post-comms-layer-legality.md, 4,910 words)

VERDICT: Partial improvement, not a solution. Sidesteps 2 of 6 broker-defining activities (lists, advertises — the seller does those themselves now). Does NOT clear the remaining four, and Ohio's Supreme Court (Dundics v. Eric Petroleum, 2018) applies the broker definition with plain-language breadth + zero exceptions.

The decisive line — "tool" vs "broker activity" — is narrower than Agent9's product vision assumes:

The OAR (Ohio Association of REALTORS®) published guidance on unlicensed assistants confirms: even discussing asking price, answering property questions, or interpreting documents without a license is prohibited. AI inherits the same statutory line.

THE 9-MINUTE CHATBOT ITSELF creates independent exposure if it:

Safe chatbot version: UI tutorial + educational scaffolding + price-range TEMPLATES (not specific recommendations). Risky chatbot version: price recommendation + AI-generated description text.

MARKET EMPIRICAL ANSWER (the closer): Every active AI-FSBO entrant that touches buyer-seller communications operates under broker license or with licensed pros embedded. REDBO explicitly states its negotiation and offer assistance is "handled by a licensed real estate agent/brokerage." FSBO.com (relaunched March 2026), Houzeo (BRKP.2022001015), Beycome (REC.2025001306) — same pattern. None thread the needle without licensed-broker coverage.

THE PATH THAT PRESERVES OWNER'S VISION: Broker-of-Record arrangement. This is REDBO's model exactly. A licensed Ohio broker whose license covers Agent9's activities; Agent9 operates all AI functionality underneath. Costs $50-$150/transaction in broker fees. Agent9 does NOT need its own license. Preserves full product, full AI functionality, full vision.

Attorney memo scope now three scenarios (original ORC 4735, MLS-skip-via-broker, seller-self-post + comms-layer) + onboarding chatbot risk + minimum structure. Estimate: $1,200-$2,000. Best-fit Ohio firms: Bricker Graydon (Columbus), Frost Brown Todd (Columbus/Cincinnati), KJK — all have published depth on ORC 4735.

UNO honest gaps: ForSaleByOwner.com's Ohio license status not verified via OREC eLicense lookup (should confirm before citing as no-license precedent); no published Ohio enforcement action against AI comms platforms (absence ≠ clearance, just novelty); no Ohio case law directly on "AI tool vs negotiation assistant" — attorney memo is the only path to certainty.

Conclusion of legal-research thread

Three rounds of research (R2-A, R2.5, R2.6) all converged on the same answer: Broker-of-record partnership is the path. Agent9 does NOT need its own broker license — but a licensed broker MUST be in the chain for ORC 4735 coverage. This is the actual industry pattern, validated by every successful FSBO/PROPTECH entrant.

The two scenarios that don't work: (a) MLS-skip alone, (b) seller-self-post + AI-comms-layer alone (without broker-of-record). The scenario that works: broker-of-record covering Agent9's AI activities. Attorney memo confirms exact activity boundaries.

Persistence discipline

Honest open questions